bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Phenological matching drives wheat pest range shift under climate change
Yuqing Wu1,2†, Zhongjun Gong1,2†, Daniel P. Bebber3*, Jin Miao1,2†, Zhonghua Zhao4, Yuying Jiang4, Shihe Xiao5, Guoyan Zhang6, Dazhao Yu7, Jichao Fang8, Xinmin Lu9, Chaoliang Lei9, Jianqing Ding10, Qiang Wang11, Yueli Jiang1,2, Tong Li1,2, Hongmei Lian1,2, Huiling Li1,2, Yun Duan1,2, Jianrong Huang1,2, Donglin Jing12, Yunzhuan He13, Zhi Zhang14, Yunhui Zhang15, Julian Chen15, Hongbo Qiao16, Wenjiang Huang17
1Institute of Plant Protection, Henan Academy of Agricultural Sciences, Zhengzhou, 450002, CN; 2Key Laboratory of Integrated Pest Management on Crops in Southern Region of North China, Zhengzhou, 450002, CN; 3Department of Biosciences, University of Exeter, Exeter EX4 4QD, UK; 4The National Agro-Tech Extension and Service Center, Beijing, 100125, CN; 5Institute of Crop Science, Chinese Academy of Agricultural Sciences, Beijing, 100081, CN; 6Henan Plant Protection and Plant Quarantine Station, Zhengzhou, 450002, CN; 7Hubei Academy of Agricultural Sciences, Wuhan, 430064, CN; 8Institute of Plant Protection, Jiangsu Academy of Agricultural Sciences, Nanjing, 210014, CN; 9Plant Sciences College, Huazhong Agricultural University, Wuhan, Hubei, CN; 430070, CN; 10Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, 430074, CN; 11Xinyang Academy of Agricultural Sciences, Xinyang, 464000, CN; 12Xingtai Institute of Agricultural Sciences, Xingtai, 054001, CN; 13College of Plant Protection, Hebei Agricultural University, Baoding, 071001, CN; 14Beijing Plant Protection Station, Beijing, 100029, CN; 15Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing, 100193, CN; 16College of Information & Management Science, Hebei Agricultural University, 450002, CN; 17Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, Beijing, 100094, CN. *Correspondence to: [email protected] † These authors contributed equally to this work.
Abstract Shifting geographical ranges of crop pests and pathogens in response to climate change pose a threat to food security (1, 2). The orange wheat blossom midge (Sitodiplosis mosellana Géhin) is responsible for significant yield losses in China (3), the world’s largest wheat producer. Here we report that rising temperatures in the North China Plain have resulted in a mean northward range shift of 3.3° (58.8 km per decade) from the 1950s to 2010s, which accelerated to 91.3 km per decade after 1985 when the highly toxic pesticide hexachlorocyclohexane (HCH) was banned (4). Phenological matching between wheat midge adult emergence and wheat heading in this new expanded range has resulted in greater damage to wheat production. Around $286.5 million worth of insecticides were applied to around 19 million hectares in an attempt to minimize wheat midge damage to crops between 1985 and 2016. Despite use of these pesticides, wheat midge caused losses of greater than 0.95 million metric tons of grain during this period. Our results demonstrate the potential for indirect negative impacts of climate change on crop production and food security, and constitute the first large scale example of plant pest range shift due to global warming.
1 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Latitudinal range shifts of wild species populations are among the strongest biological
signals of recent anthropogenic climate change (5). Similar range shifts in invasive
species or pests and pathogens of agricultural crops and livestock would have significant
implications for biodiversity conservation or food production, respectively. While there is
some evidence for latitudinal range shifts in line with global warming for crop pests and
pathogens in general (1), few emerging pest outbreaks have been unequivocally
attributed to climate change (2). Similarly, while many studies have modelled the effects
of climatic change and variability on crop production, the impact of pests and diseases is
generally ignored. Given the huge production losses attributable to crop pests and
pathogens (6), and the economic and environmental burdens resulting from their control,
understanding how climate change alters the risk of outbreaks is key to improving global
food security in a warming world.
Wheat (Triticum aestivum) is the world’s most important crop in terms of production
area and second-most important in human diets. Wheat is grown mainly at temperate
latitudes in both hemispheres, with China the world’s largest producer. The North China
Plain (NCP, Fig. 1, Supplementary Fig. S1), comprising the provinces of Beijing, Tianjin,
Hebei, Shandong, Henan, Anhui, Jiangsu, Shanghai, Zhejiang, Jiangxi and Hubei, grew
around one tenth of the world’s wheat between 1961 and 2016. Orange Wheat Blossom
Midge (hereafter, wheat midge), Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae)
is a major wheat pest in northern latitudes around the world, including in the NCP (3).
Wheat midge is univoltine and has a long diapause period during which cocooned larvae
overwinter in the soil (7). Diapause is broken by winter chilling (7), followed by pupation
and emergence of adults from the soil in response to warming and other climatic cues (8–
10). In North America and Europe, wheat midge larvae consume the kernels of spring
wheat and other Poaceae crops, resulting in serious losses. An outbreak in northeast
Saskatchewan, Canada in 1983 caused losses of $30 million across nine municipalities
(11). Since this time, wheat midge has spread west and south and is now present in many
regions of the Great Plains of North America (12). Adult midges are able to disperse
many hundreds of kilometers on the wind (13). In China, one of the first reported 2 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
outbreaks occurred in Wuxian, Jiangsu province in 1838 (14). By the 1950s, wheat midge
had spread into the winter wheat production area in the southern region of the NCP
between 28° N and 36° N (15). Henan province, for example, suffered wheat production
losses of 0.352 million tons (7.7 % of production) from an outbreak in 1954 (15).
The history of wheat midge outbreaks in China has been strongly influenced by the
pesticide 1,2,3,4,5,6-Hexachlorocyclohexane (HCH), also called benzene hexachloride
(BHC) or lindane (4). HCH was introduced as a soil treatment in the 1950s, keeping
outbreaks in the NCP under control over the period 1958 to 1982 (Fig. 2). Production of
HCH in China grew steadily to the mid-1970s, peaking around 250 kt before dropping to
zero in 1984 (4). The NCP provinces of Jiangsu, Hunan, Henan, Shandong, Hubei,
Zhejiang, Jiangxi and Hebei were among the top ten provinces with the highest HCH
usage between 1952 and 1984 (4). However, serious environmental problems lead to a
ban on the application of HCH in 1983 (4), resulting in the resurgence in wheat midge
and a cumulative loss in wheat production of 2.77 million tons between 1985 and 2016
(Fig. 2). The resurgence of wheat midge was accompanied by a shift in the insect’s
geographical range. Given predicted latitudinal range shifts under future climate change
for this pest (16), and historical range shifts observed for crop pests and pathogens in
general (2), the altered wheat midge distribution across the NCP over recent decades
could be a result of phenological responses to climate change. Here, we investigate the
pattern and causes of this range shift, testing the hypotheses that warming
temperatures due to climate change allowed the insect to invade the northern regions of
the NCP, and that climate change altered phenological matching that is key to successful
host exploitation by the pest.
Wheat midge was restricted to the southern half of the NCP in the 1950s, and the
northern half in the 2010s (Fig. 1). Wheat midge damage records from 135 Counties
during the period 1950 to 1959 were distributed over the provinces of Henan, Anhui,
Jiangsu, Hubei, Jiangxi, Zhejiang and Shanghai across a latitudinal range of 28° N to 36°
N. We term this region the Original Distribution Range (ODR). In comparison, 225 3 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
reports of wheat midge damage in the NCP were recorded in the 2010s, across eight
provinces (Hebei, Beijing, Tianjin, Shandong, Henan, Anhui, Jiangsu and Hubei)
between 32° N and 40° N. Four of these provinces, Hebei, Shandong, Beijing and Tianjin,
comprise a Novel Expanded Range (NER) in the northern part of the NCP that had no
outbreaks during the 1950s. There were no reports of wheat midge damage in three
southern provinces (Jiangxi, Zhejiang and Shanghai) below latitude 30° N in the 2010s,
each of which that had recorded pest damage prior to 1960. The latitudinal median for
1950s observations was 32.74 °N (IQR 32.02 – 33.60 ° N), compared with 35.81 °N (IQR
33.81 – 37.95 °N) for the 2010s. The median latitude of midge outbreaks was 3.21 ± 0.02°
SE further north in the 2010s than in the 1950s, equivalent to a range shift of 58.8 ± 0.4
km per decade (bootstrap mean of difference between median outbreak latitude in 2010s
vs. 1950s). First observations of wheat midge larvae in soils at 37 sites across the NER
(Supplementary Fig. S2) revealed that the insect spread northwards at a rate of 0.083 ±
0.021° N y-1, or 91.3 ± 23.3 km per decade since the resurgence of wheat midge in 1985
(Fig. 3). These rates of migration closely match climate velocities for the region since
1960 (17), and are within the observed range for dipteran crop pests in general (1).
Poleward range shifts are expected under global warming (2) but we must discount any
effect of changing host distribution before attributing climate change as a cause of wheat
midge range shift. In the NCP, the southernmost provinces of Jiangxi and Zheijang have
historically grown little wheat, declining to < 1 % of total area by the 2010s
(Supplementary Fig. S3). Hence, the disappearance of wheat midge from Jiangxi and
Zhejiang could be due to absence of the host. Shanghai, however, has retained 7.5 %
wheat cover (down from 10.8 % in the 1950s) but did not record an outbreak in the 2010s.
The central NCP provinces of Hubei, Henan, Anhui, and Jiangsu have retained moderate
to high wheat cover (mean 5.9, 27.0, 14.9 and 18.2 %, respectively) with little change over
time. Hebei and Shandong, which make up the majority of the NER, have relatively high
and stable wheat area (12.7 and 24.3 %, respectively), while the smaller regions of
Beijing and Tianjin have relatively low cover with only Tianjin showing an increase since
the 1950s. Thus, the NER has always grown significant areas of wheat, and the 4 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
northward range expansion of wheat midge cannot be attributed to host distribution
shifts.
Wheat grain consumption by wheat midge larvae is dependent upon oviposition during
wheat heading in spring (3). Mean spring temperatures (February, March, May
temperatures from CRU gridded estimates) increased by 0.035 ± 0.007 ºC y-1 across the
NCP between 1951 and 2017, increasing more rapidly in the NER (0.045 ± 0.08 ºC y-1)
than the ODR (0.031 ± 0.06 ºC y-1, Supplementary Fig. S4). Median spring temperatures
in the 1950s for sites that experienced outbreaks in the 1950s were 8.8 ºC (IQR 8.4 – 9.3
ºC), whereas sites that were to experience outbreaks in the 2010s were cooler with
median temperature 7.5 ºC (IQR 6.6 – 8.4, Fig. 4). By the 2010s, median spring
temperatures had risen to 9.3 ºC (IQR 8.5 – 10.1 ºC) for those sites experiencing
outbreaks in the 2010s, while those which had experienced outbreaks in the 1950s had
warmed to median 10.3 ºC (IQR 9.8 – 10.8 ºC). Thus, outbreaks are now occurring at
temperatures around 0.5 ºC warmer than in the 1950s. The median spring temperature
during the year of arrival at each of 37 monitoring sites in the NER was 8.5 ºC (IQR 7.5 –
8.9 ºC).
Wheat midge larvae consume developing wheat kernels, and adult oviposition must
coincide with wheat heading, since the susceptibility of plants to wheat midge damage
declines significantly after anthesis begins (18, 19). Wheat and midge phenology were
monitored in five sites from 30 to 39 °N in 2013, 2014 and 2016, to quantify the degree of
phenological matching (Supplementary Figs. S2 & S5, Supplementary Table S1). Peak
wheat heading occurred between 3rd April and 6th May depending on year and site, while
the date by which at least half of midge adults had emerged varied between 16th March
and 11th May. Wheat heading and adult midge emergence were delayed along increasing
latitudes, by 2.93 d ± 0.37 SE per degree latitude for wheat, and 4.39 d ± 0.51 SE per
degree latitude for wheat midge (estimated by linear mixed effects models with random
intercepts per site). Positive mismatches (when wheat midge adults emerged before
wheat heading) were observed at warmer lower latitudes, and negative mismatches 5 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
(wheat heading before wheat midge emergence) at cooler, higher latitudes (Fig. 5a,b).
Zero mismatch days occurred at spring temperatures of 10.1°C ± 0.4 SE (Fig. 5b). The
number of larvae found in soil increased significantly with latitude (Fig. 5c, correlation r
= 0.80, df = 13, p = 0.0003). The largest numbers of larvae per wheat ear were found at
mismatches of -3 to 0 days, with no larvae present when phenologies were strongly
mismatched (Fig. 5d). Thus, wheat midge damage is greatest at higher latitudes, where
adult emergence is better synchronized with wheat heading.
We estimated the mismatch between wheat heading and adult midge emergence across
the NCP in the 1950s and 2010s, from a linear regression of mismatch days on spring
temperature at the five observation sites (Fig. 5b). We applied the regression equation
(mismatch days = -20.4 + 2.07 ´ spring temperature) to spring temperatures derived
from monthly CRU temperature data (Fig. 6). The estimated mismatch days correlate
with the observed distribution of midge outbreaks in the 1950s and 2010s (Fig. 1).
Slightly negative mismatches, which are associated with the greatest infestations (Fig.
5d), occur in the ODR provinces of Henan, Anhui and Jiangsu during the 1950s (Fig. 6a).
By the 2010s, areas of zero to slightly negative mismatches had shifted northward into
the NER provinces of Hebei and Shandong (Fig. 6b). Temperature-driven phenological
patterns detected in recent observational studies therefore strongly predict the locations
of historical outbreaks.
Emergence of wheat midge adults has been modelled using thermal time (cumulative
degree days above a threshold temperature)(8, 20) or via temperature thresholds (9, 21).
At our six phenology monitoring sites, adult wheat midge emergence peaked on days
with a median temperature of 17.7 °C (IQR 16.6 – 19.1 °C), wheat heading peak had a
much wider temperature range (median 17.7 °C, IQR 14.1 – 20.5 °C, Supplementary Fig.
S6). Conversely, wheat heading occurred over a relatively small range of degree days
above 5°C (median 438.5, IQR 398.5 – 460.5), compared with adult wheat midge
emergence (median 447.8, IQR 391.4 – 481.9). Hence, a temperature threshold may be a
more reliable predictor than temperature accumulation for adult wheat midge emergence 6 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
(9). While some studies have suggested precipitation events as triggers for midge
emergence (9, 10), we found no suggestion of a relationship between rainfall and
emergence (Fig. S5).
The northward range shift of wheat midge has had significant effects on food production
and pest control expenditure. National Agro-tech Extension and Service Centre of China
(NATESC) records show that 19.07 million ha of winter wheat was treated with
insecticides in an attempt to minimize wheat midge damage in the NERC between 1985
and 2016 (Fig. 2). The cost of these insecticides (approximately US $15 per ha) means
that climatic warming has caused an estimated economic loss of around $286.05 million
to Chinese wheat producers during this period. Furthermore, increased insecticide usage
has only been partially successful, with around 0.95 million metric tons of wheat still
being lost to wheat midge damage from 1985 to 2016 (Fig. 2). Our results show that
climate warming caused a northward shift in wheat midge outbreaks in the NCP. This
geographical shift has resulted in the peak date of midge adult emergence coinciding
more closely with winter wheat ear heading dates at higher latitudes. Wheat midge is
also anticipated to expand its range in the Northern Great Plains of North America as
temperatures increase, which is likely to affect spring wheat harvests over the next 40
years (16). There is growing interest in the use of parasitoids for biological control (22),
acknowledging the importance of phenological matching between the parasitoid and the
midge to maximize efficacy (10, 23). However, little is known of the responses of these
natural enemies to climate change, nor whether biocontrol could surplant chemical
controls in future.
Over 80 arthropods, pathogenic microorganisms and nematode taxa could potentially
cause damage to wheat species in China (24). The range of many of these other pest
species has also expanded into northern winter wheat production areas in recent decades,
with heavy pesticide use required for their control. For example, the fungal pathogen
Fusarium head blight (Fusarium graminearum) was first detected in Henan province in
2012 (25). Despite the importance of these pests and diseases, their impacts, monitoring 7 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
and management have often been ignored in considerations of yield improvements for
wheat in China (26, 27). The rapid invasion of major wheat growing regions in the NER
highlights the urgency with which climate change impacts on agricultural pests and
pathogens must be addressed.
METHODS
The study spans the North China Plain (NCP), a low-lying plain ranging from 111° E to
123° E and 28° N to 40° N, and covering a total area of around 400,000 km2 (Fig. 1,
Supplementary Fig. S1). The NCP comprises eleven provinces, of which seven (Henan,
Anhui, Jiangsu, Hubei, Jiangxi, Zhejiang and Shanghai) make up our Original
Distribution Range (ODR) for wheat midge, and four (Hebei, Shandong, Beijing and
Tianjin) make up the Novel Expanded Range (NER).
We obtained agricultural and meteorological data for the NCP from a number of sources.
Data for annual wheat production and harvested areas of each province were obtained
from the crop database of the Ministry of Agricultural and Rural Affairs of the People’s
Republic of China (http://202.127.42.157/moazzys/nongqing.aspx). Gridded wheat
production area estimates at 10km spatial resolution for the year 2005 (used in Fig. 1)
were obtained from MAPSPAM (http://mapspam.info). Monthly mean temperature data
at 0.5° spatial resolution from January 1951 to December 2017 were obtained from the
CRU TS v. 4.02 reanalysis (https://crudata.uea.ac.uk/cru/data/hrg/cru_ts_4.02/). Spring
was defined as the months February, March and April (1 February to 30 April). For
specific research sites, temperature and precipitation data were obtained from the China
National Meteorological Administration (http://eng.nmc.cn/).
Analyses of wheat midge distribution and phenology were based on three datasets: an
NCP-wide compilation of wheat midge outbreaks in the 1950s and 2010s; measurements
of wheat midge soil larval densities, larvae per wheat ear and wheat grain loss at 37
experimental stations from 1985 onwards; and measurements of soil larval densities,
8 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
adult emergence phenology, and wheat phenology at five locations across a latitudinal
gradient in 2013, 2014 and 2016.
Records of location, earliest dates of 135 counties recorded damage of larvae feeding on
wheat kernel in the 1950s were obtained from the Institute of Plant Protection and Soil
Science, Hubei Academy of Agricultural Sciences (formerly known as the Central South
Institute of Agricultural Sciences), Institute of Plant Protection, Jiangsu Academy of
Agricultural Sciences (formerly known as the East China Institute of Agricultural
Sciences), and from the literature (14, 15, 28). These records originate from seven
provinces (Henan, Anhui, Jiangsu, Hubei, Jiangxi, Zhejiang and Shanghai). Records of
225 counties on wheat midge incidence since 2010 were obtained with permission from
NATESC (https://www.natesc.org.cn/).
From 1985 to 2014, soil larval densities were sampled in 37 wheat experimental stations
of the China Agricultural Research System (CARS), Plant Protection Plant and Plant
Quarantine Station in Henan province, and Beijing City, giving coverage across the NER
between 36° N and 40° N (Fig. S7). Soil samples (10×10×20 cm, length width and depth)
were taken after annual harvesting of wheat in September. This involved diagonal
sampling of 5 points in a field, with 10 randomly-selected fields per county (People’s
Republic of China National Agricultural Standard NY/T616-2002, 2003). At the same
sites, the fraction of wheat yield loss at each wheat grain filling stage was estimated from
the number of larvae per ear (People’s Republic of China National Agricultural Standard
GB/T 24501-2009). Measurements were carried out in more than 10 representative
wheat fields in each of the 37 sites used for soil larval density measurements. Diagonal
5-point sampling was carried out in each field, with 10 ears of wheat collected per
sampling point, and each ear checked for number of larvae and the number of grains.
Percentage wheat yield losses L = 100´W/GC where W is the total number of larvae per
ear, G is the total number of grains per ear, and C is the number of larvae required to
consume a grain (C = 4).
9 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Dates for wheat heading and flowering periods were recorded from March, with plots
covered with 100-mesh gauze and the number of adult insect eclosions recorded daily.
Close synchronization between the heading stages of winter wheat and the breeding
cycle of wheat midge adults would lead to serious losses in wheat yields. Therefore,
synchronization of the wheat heading period and the peak of adult insect emergence were
determined for six different geographical sites along a latitudinal gradient: Wuhan
(30.36 °N), Xinyang (32.08 °N), Xinxiang (35.19 °N), Xingtai (37.04 °N), Baoding
(38.44 °N) and Langfang (39.30 °N) (Supplementary Fig. S5, S7, Supplementary Table
S1). Boading was monitored in 2013 and 2014, after which monitoring moved to
Langfang in 2016. Three main winter wheat varieties (Bima 1, Xiaoyan 4 and Zhoumai
18) were used for wheat production in these regions, with wheat damage levels
determined by monitoring the number of larvae present per 100 ears in the different
regions in 2013, 2014 and 2016.
Data availability
Locations of all auxiliary datasets are listed in the Methods.
All experimental datasets will be made available on DRYAD (https://datadryad.org) upon
manuscript publication.
10 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
References
1. D. P. Bebber, M. A. T. Ramotowski, S. J. Gurr, Crop pests and pathogens move polewards in a warming world. Nature Clim. Change. 3, 985–988 (2013).
2. D. P. Bebber, Range-Expanding Pests and Pathogens in a Warming World. Annu. Rev. Phytopathol. 53, 335–356 (2015).
3. Y. Duan et al., Genetic Diversity and Population Structure of Sitodiplosis mosellana in Northern China. PLOS ONE. 8, e78415 (2013).
4. Y. F. Li, D. J. Cai, Z. J. Shan, Z. L. Zhu, Gridded Usage Inventories of Technical Hexachlorocyclohexane and Lindane for China with 1/6° Latitude by 1/4° Longitude Resolution. Arch. Environ. Contam. Toxicol. 41, 261–266 (2001).
5. IPCC, Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, Cambridge, UK, 2014).
6. M. C. Fisher et al., Emerging fungal threats to animal, plant and ecosystem health. Nature. 484, 186–194 (2012).
7. W. Cheng, Z. Long, Y. Zhang, T. Liang, K. Zhu-Salzman, Effects of temperature, soil moisture and photoperiod on diapause termination and post-diapause development of the wheat blossom midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae). Journal of Insect Physiology. 103, 78–85 (2017).
8. R. H. Elliott, L. Mann, O. Olfert, Calendar and degree-day requirements for emergence of adult wheat midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae) in Saskatchewan, Canada. Crop Protection. 28, 588–594 (2009).
9. G. Jacquemin, S. Chavalle, M. De Proft, Forecasting the emergence of the adult orange wheat blossom midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae) in Belgium. Crop Protection. 58, 6–13 (2014).
10. S. Chavalle, P. N. Buhl, F. Censier, M. De Proft, Comparative emergence phenology of the orange wheat blossom midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae) and its parasitoids (Hymenoptera: Pteromalidae and Platygastridae) under controlled conditions. Crop Protection. 76, 114–120 (2015).
11. O. O. Olfert, M. K. Mukerji, J. F. Doane, Relationship between infestation levels and yield loss caused by wheat midge, Sitodiplosis Mosellana (Géhin) (Diptera: Cecidomyiidae), in spring wheat in Saskatchewan. The Canadian Entomologist. 117, 593–598 (1985).
11 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
12. O. Olfert, R. H. Elliott, S. Hartley, Non-native insects in agriculture: strategies to manage the economic and environmental impact of wheat midge, Sitodiplosis mosellana, in Saskatchewan. Biol Invasions. 11, 127–133 (2009).
13. J. Miao et al., Long-Distance Wind-Borne Dispersal of Sitodiplosis mosellana Géhin (Diptera:Cecidomyiidae) in Northern China. J Insect Behav. 26, 120–129 (2013).
14. X. Zeng, Wheat Midge (Agricultural Press, Beijing, 1965).
15. P. L. Yang, Research and control on Wheat Midge (Science Press, Beijing, 1959).
16. O. Olfert, R. M. Weiss, R. H. Elliott, Bioclimatic approach to assessing the potential impact of climate change on wheat midge (Diptera: Cecidomyiidae) in North America. The Canadian Entomologist. 148, 52–67 (2016).
17. M. T. Burrows et al., The pace of shifting climate in marine and terrestrial ecosystems. Science. 334, 652–655 (2011).
18. Y. Wu et al., The synchronization of ear emerging stages of winter wheat with occurrent periods of the orange wheat blossom midge, Sitodiplosismosellana (Gehin) (Diptera: Cecidomyiidae) adults and its damaged level. Acta Ecologica Sinica. 35, 3548–3554 (2015).
19. R. H. Elliott, L. W. Mann, Susceptibility of red spring wheat, Triticum aestivum L. cv. Katepwa, during heading and anthesis to damage by wheat midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae). The Canadian Entomologist. 128, 367– 375 (1996).
20. J. F. Doane, O. Olfert, Seasonal development of wheat midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae), in Saskatchewan, Canada. Crop Protection. 27, 951–958 (2008).
21. J. N. Oakley et al., Prediction of orange wheat blossom midge activity and risk of damage. Crop Protection. 17, 145–149 (1998).
22. S. Chavalle, P. N. Buhl, G. San Martin y Gomez, M. De Proft, Parasitism rates and parasitoid complexes of the wheat midges, Sitodiplosis mosellana, Contarinia tritici and Haplodiplosis marginata. BioControl. 63, 641–653 (2018).
23. R. H. Elliott, L. Mann, O. Olfert, Calendar and degree-day requirements for emergence of adult Macroglenes penetrans (Kirby), an egg-larval parasitoid of wheat midge, Sitodiplosis mosellana (Géhin). Crop Protection. 30, 405–411 (2011).
24. Y. Y. Guo, Crop diseases and insect pests (China Agricultural Press, Beijing, 2015).
25. F. Xu et al., First Report of Fusarium pseudograminearum from Wheat Heads with Fusarium Head Blight in North China Plain. Plant Disease. 99, 156–156 (2014).
12 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
26. X. Qin et al., Wheat yield improvements in China: Past trends and future directions. Field Crops Research. 177, 117–124 (2015).
27. S. Sun, X. Yang, X. Lin, G. F. Sassenrath, K. Li, Climate-smart management can further improve winter wheat yield in China. Agricultural Systems. 162, 10–18 (2018).
28. C. C. Long, White midge S. mosellana in Jianxi Province. Bulletin of Agricultural Sciences. 3, 147 (1955).
Acknowledgements
YW was supported by the China Agricultural Research System (CARS-03) and Natural
Science Foundation of Henan (182300410003). DB was funded by the Agri-tech Cornwall
and Isles of Scilly programme, supported by the European Regional Development Fund,
Cornwall Council, the Council for the Isles of Scilly, the Cornwall College Group, the
Universities of Exeter and Plymouth and Rothamsted Research.
Author contributions
YW, GZ, HZ, SX, GZ, DY J F, ZZ, and DY collected the data. TL, JH, HL, YJ, YJ, JD, CL,
QW, DJ YZ, and JC performed the experiments. YW, ZG, DPB, JM & XL analysed the
data. YW, DPB, ZG, JM and XL wrote the paper.
Competing financial interests
The authors declare no competing financial interests.
13 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
40 80 BJ 60 HE TJ
38 40 20 SD 0 ● 36 ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ●● ● ●● HA ● ● ● ●● ● JS N 34 ● ● ● ● ● ● ● ● ° ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ●● ●● ● ● ● ●● ● ● 32 ● ● ● ● ● ● ● ● ● HB ● ● AN ● ● ● ● SH ● ● ●
30 ● ● ●● ZH JX ● ● ● ● ● ● ● 0 200 ● ●● ● ● ● 28 km ●
112 114 116 118 120 122 ° E Figure 1. Wheat midge outbreaks in the NCP.
Blue circles show outbreaks recorded in the 1950s in the Original Distribution Range
(ODR). Red triangles show wheat midge outbreaks from 2010 in the Novel Expanded
Range (NER). The NCP is in heavy outline. The grey scale represents the areal wheat
planting density in 2005, from MAPSPAM estimates. Provinces in the ODR are Henan
(HA), Jiangsu (JS), Hubei (HB), Anhui (AH), Shanghai (SH), Jiangxi (JX) and Zhejiang
(ZH). Provinces in the NER are Beijing (BJ), Tianjin (TJ), Hebei (HE) and Shandong (SD).
Sea is pale blue.
14 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
a b Region 8 ODR HCH NER 10 6
4 8
Wheat (Mha) Wheat 2 non-HCH Insecticide (Mha) Insecticide 6 0 1960 1980 2000 1960 1980 2000 c d 0.20
2.0 0.15
1.5 0.10 1.0 Loss (Mt) Loss 0.05 0.5 Wheat midge (Mha) midge Wheat
0.0 0.00 1960 1980 2000 1960 1980 2000 Figure 2. Wheat and the wheat midge in the NCP, 1950 – 2016.
a. Wheat production area in the Original Distribution Range (ODR, blue line) and New
Expanded Range (NER, red line) provinces (Mha). Grey vertical line at 1985 highlights
ban on HCH application. b. Insecticide-treated area (Mha), red and blue show
HCH-treated area, pink and pale blue show non-HCH insecticide treated area. c.
Wheat-midge affected area (Mha). d. Wheat production loss (Mt).
15 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
40 ) N
° 39 (
38 Latitude 37
1990 2000 2010 Year Figure 3. Wheat midge first arrival across the NER.
Latitude of the first year in which wheat midge larvae were detected in soil in the NER,
1985-2014. Mean latitude of first observation increased by 0.083 ± 0.021 ° y-1 (linear
regression, n = 37, r2 = 0.31, p < 0.001).
16 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
1950s 2010s ● ● ● ● ) ● ● ● C ●
° ●
( ● 10
● ● ● 5 ● ● ● ● Temperature Temperature ● ● NER ODR NER ODR Region Figure 4. Spring temperatures of locations of wheat midge outbreaks in the
ODR and NER in the 1950s and 2010s.
17 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
a b ● 2013 2014 10 2016 10 ● ●
0 0 ● ● Mismatch (d) Mismatch ● ● (d) Mismatch ● ● ● ● −10 30.0 32.5 35.0 37.5 7 9 11 13 Latitude (°N) Spring temperature (°C)
c d ● ● ● ● 3 ● 3 1 − soil 1 − 2 2
● 1 1 Larvaeear Larvael ● ● ● 0 0 ●
30.0 32.5 35.0 37.5 0 10 Latitude (°N) Mismatch (d) Figure 5. Wheat and wheat midge phenology.
a) Mismatch between date of wheat midge adult emergence and date of wheat ear
heading in the NCP vs. latitude, in 2013 (blue circle), 2014 (red triangle) and 2016 (green
square), with line of best fit (grey). Positive mismatch indicates peak wheat midge
emergence occurs before peak wheat heading, negative values that peak wheat midge
emergence occurs after peak wheat heading. b) Mismatch vs. spring temperature. c)
Larvae per litre of soil vs. latitude. d) Mean number of wheat midge larvae per wheat ear
vs. mismatch days.
18 bioRxiv preprint doi: https://doi.org/10.1101/614743; this version posted April 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
a b
10 5 0 −5 −10
Fig. 6. Estimated phenological mismatch between wheat heading and adult
midge emergence (days) in a) the 1950s and b) 2010s.
Mismatch days were estimated from mean spring temperatures (monthly CRU data) by
linear regression of observational data (see text for details). Grey regions lie outside the
spring temperature range in the observational study (-5 to +13 °C).
19